Blood samples were obtained from 181, Aberdeen Angus (AA, n=98) and Charolais (C, n=83), multiparous cows, in the postpartum period. The animals came from a previous study which compared the efficiency of different hormonal treatments associated with 96-hour calf removal in relation to complete weaning of animals fed with different forages (TERRA, 2001). The animals were adult cows (ages varying from 4 to 6 years), with mean body condition at delivery of 3.0 (in a classification range from 1 - very thin - to 5 - obese; LOWMAN et al., 1976). Fifty to 70 days postpartum, the cows previously classified at delivery, were ranked according to their body condition and submitted to hormonal treatment. All animals were weighed twice (at delivery and at weaning - performed 7 days after start of treatment) and sorted in two groups on the basis of forage availability: group one was managed on native pasture with 960kg of dry matter per hectare (DM ha-1) and a stocking rate of 0.96 animal unit per hectare (au ha-1; au=400kg live weight) at delivery, and 400kg DM ha-1 at weaning; Group 2 was also managed on native pasture but with 600kg DM ha-1 and a stocking rate of 1.44au ha-1 at delivery and 240kg DM ha-1 at weaning. The dry matter of the pasture was estimated by the double sample method (WILM et al., 1944). Some animals gained while others lost weight in both groups (Table 1) suggesting that factors other than forage availability (which was above the maintenance diet, estimated based on dry matter only) must have influenced daily weight gain.
All cows that showed oestrous between days 7 and 17 from the beginning of the treatment were artificially inseminated; they were then bred with a cow: bull ratio of 100:12 up to day 67; clinical and ultrasonic diagnosis of pregnancy were performed at day 60 from the beginning of experiment to calculate the proportion of cows that conceived to the first oestrous after treatment and at day 127 to estimate the final pregnancy rate.
Blood was obtained from the jugular vein using ACD (acid-citrate-dextrose) as anticoagulant (ALMEIDA et al., 2003). Genomic DNA was extracted from peripheral blood leukocytes by the method of MILLER et al. (1988). Short tandem repeats were PCR-amplified as indicated in STEIGLEDER et al. (2004) and SNPs according to POMP et al. (1997) and BUCHANAN et al. (2002). The STR amplification products were analyzed by vertical electrophoresis in non-denaturing polyacrylamide gel (LAHIRI et al., 1997) and those of SNPs in agarose gel after Sau3AI or Kpn2I endonucleases cleavage.
Variance analyses (one-way ANOVA) were performed to compare productive parameters [average daily weight gain (ADG) and pregnancy according to the diagnosis at days 60 (P1) and 127 (P2)] in the different genotype classes according to the model: Yij = µ + Ai + eij
Where, Yij is the ADG, P1 or P2 phenotype of the jth individual; µ is the effect of the population mean; Ai is the effect of the ith genotype class; and eij is the random error component. The Tukey post hoc test was employed to identify the significant group, when a significant association was detected. Descriptive statistics was carried out beforehand to verify the normality of the distribution of productive parameters.
To compare the significant group detected by ANOVA with the other genotypes, average daily weight gain was also dichotomized into classes consisting of animals heavier or lighter than the mean ADG population values (see Table 1). For the analysis considering both populations simultaneously, the mean value was corrected by the herd weight; this analysis was made only when the positive association was detected in both populations. The genotype frequencies were compared between heavier and lighter animals by the chi-square test. To estimate the effect of the genetic markers on ADG, the odds ratio (OR) with 95% confidence interval was obtained by univariate logistic regression analysis. All the statistic analyses were performed using the SPSS® for WindowsTM software (SPSS Inc), version 10.0.5 (1999).